The assumption that metformin — a proven insulin-sensitizing agent — could intercept gestational diabetes before it develops has guided multiple clinical trials over the past decade. This patient-level meta-analysis, published in NEJM Evidence, delivers a sobering corrective: for women at elevated metabolic risk, prophylactic metformin does not meaningfully reduce the likelihood of developing gestational diabetes mellitus (GDM).

Drawing on individual participant data (IPD) from seven of ten eligible double-blind, randomized placebo-controlled trials — representing 2,297 pregnancies after harmonization — researchers applied one-stage mixed-effects models adjusting for maternal age, BMI, gestational age at commencement, and baseline glucose. Across the two most widely used diagnostic frameworks — WHO 1999 criteria (adjusted OR 1.00; 95% CI 0.71–1.41) and NICE 2015 criteria (adjusted OR 1.00; 95% CI 0.71–1.41) — metformin showed no association with GDM reduction. A marginal signal emerged under IADPSG thresholds on adjusted analysis, though this did not constitute consistent evidence of benefit. Primary neonatal outcomes, including gestational age at delivery and anthropometric measures, showed no compelling metformin advantage.

This finding is significant precisely because IPD meta-analyses represent the methodological gold standard — pooling raw participant data rather than aggregate trial statistics eliminates much of the noise that plagues conventional meta-analyses. With nearly 2,300 pregnancies analyzed, statistical power was adequate to detect clinically meaningful effects. The null result therefore carries unusual credibility. It challenges a biologically intuitive hypothesis: metformin improves insulin sensitivity in non-pregnant adults, but placental physiology and the trajectory of gestational glucose intolerance appear more complex than that extrapolation implies. Clinicians and guideline bodies evaluating metformin as a GDM prevention strategy — particularly in obese or polycystic ovary syndrome populations — should weigh this evidence carefully. The study's scope was prevention, not treatment; established GDM management data remain separate. Overall, this is a paradigm-narrowing finding that reframes metformin's role in obstetric care.